Lecture calendar
- Monday 27/05, Room 0.5, h. 12.00 - 14.00
- Wednesday 29/05, Room 0.7, h. 13.30 - 16.30
About the speaker
Eric Grivel received his PhD in signal processing in 2000 in Bordeaux (France). He joined Bordeaux Institute of Technology (Bordeaux INP) in 2001 as an assistant professor and then as a professor in 2011. For more than 20 years, he has been with the Signal & Image research group at IMS lab (which is a joint research unit for the French National Center for Scientific Research CNRS, University of Bordeaux and Bordeaux INP). His research activities deal with statistical signal processing with applications in speech and audio processing, mobile communication systems, radar processing, GPS navigation and biomedical. From 2010 to 2017, he was the head of the Telecommunications department at ENSEIRB-MATMECA, one of the graduate schools of engineering at Bordeaux INP. From 2014 to 2022, he was in charge of the Scientific Interest Group with Thales, University of Bordeaux, University of Poitiers, University of Limoges, Bordeaux INP, INRIA and CNRS. Since 2021, he is the Vice Dean of the international relations of Bordeaux INP.
Abstract
Kalman filtering has been used in a wide range of fields (control, signal processing and econometrics) and various applications (speech enhancement, channel estimation in mobile communication systems, object tracking in radar applications, car tracking in video, GPS navigation, trend estimation, on-line parameter estimation, control of vehicles and aircraft, etc.). Kalman filtering is based on a state space representation of the system. We will see how to move from the time-continuous representation to the discrete-time one. More particularly, various cases of motion model will be addressed. They are useful when dealing with object tracking and GPSnavigation. They describes how an object moves with respect to time. Position, velocity and sometimes the acceleration are considered. Thus, the constant-velocity motion model, the constant-acceleration motion model but also the Singer model will be presented. Other state space representation of the system will be considered, especially when dealing
with autoregressive processes and moving average processes which are often used to represent time series. Then, the main steps of Kalman filter will be presented. During the talk, additional comments on how to use Kalman filtering (especially when the dynamics of the system change over time) will be given.